Google’s Gemini AI chatbot has swiftly become a major player in the generative AI landscape. Despite its initial hiccups, Gemini stands out as one of the few genuine competitors to ChatGPT. Let’s delve into all you need to know about it.
Google initially unveiled its AI venture as Bard in March 2023, offering it free and experimentally. The chatbot originally ran on the LaMDA large language model (LLM). August 2023 saw the introduction of Google Duet, an enterprise option featuring AI-infused Workspace apps like Gmail, Drive, Slides, Docs, and others. However, in December 2023, Google upgraded the Bard language model to the Gemini LLM. By February, they unified Duet and Bard into a single product and rebranded all their AI offerings under the unified name: Gemini. As of June 2024, Gemini is accessible on Android in 29 languages for any device running Android 10 or later with at least 2GB of RAM. iOS users can access the service through the Gemini tab within the Google app.
Similar to ChatGPT, Google Gemini is an AI chatbot capable of generating text and answering queries. It essentially can handle anything ChatGPT can. Google is positioning its Gemini generative AI as exceptionally potent due to its multimodal capabilities, enabling it to perform tasks like generating text, images, video, audio, and code within the same platform. Gemini supports various large language model (LLM) options tailored for different tiers of purpose and expertise, including Gemini Nano, Gemini Pro, Gemini Ultra, and Gemini Advanced.
Google harbors ambitious plans to make its AI useful across a multitude of applications. Following trials of Gemini integration into its Chrome browser, Google officially implemented the technology into its Search function in May at I/O 2024. It has also been integrated into Google’s Workspace apps, including Gmail, Drive, Slides, Docs, and others. These features were announced at I/O 2024 and began rolling out to select users in June of that year. Google also provides a paid tier called Google One AI Premium, offering several benefits, including a significantly higher number of queries users can ask the most advanced version of Google’s AI (GPT-4o) within a given three-hour period.
Continue reading to discover how to get started with Google Gemini and determine which Gemini option best suits your requirements.
Signing up is straightforward if you’re looking to use the free version of Google Gemini. Many individuals already have a Google account and email. Simply use your Google credentials to access the Gemini site, and you’re good to go. If needed, you can sign up for a Google account beforehand. You might encounter terms and services notifications, which you can review and confirm at your discretion. You may then see a notice that Google Bard is now Google Gemini. After this, you’ll be presented with a page resembling most typical AI chatbot setups. It includes a personalized introduction with your name, some prompt suggestions, and a prompt text box at the bottom. In the upper-left corner is a Gemini drop-down menu where you can refresh your page for a new chat or choose to upgrade to the Gemini Advanced paid tier. The furthest left corner features a three-line icon that allows you to expand the menu. Here, you can see a new chat tab, your recent prompts, a help option, activity options, settings, and another option to upgrade to Gemini Advanced. Select the three-line icon again to collapse the menu. You’ll find your Google account icon in the upper-right corner. Select it to view your Google account details and any other connected accounts.
Google Gemini, similar to many other AI chatbots, boasts the ability to process and generate different types of text through prompts. Google utilizes examples like explaining complex topics and event planning for potential prompts. Common uses for Gemini include story and essay generation, language translation, code generation and editing, solving math equations, and more. You can begin a prompt with a strong action statement such as “write,” “solve,” “summarize,” “generate,” “create,” or “help me” to convey your request and provide as much detail as needed to ensure the best outcome.
One prompt I used was “teach me about the theory of relativity in simple terms,” and Gemini returned a detailed yet concise response about the scientific theory. It generated three drafts for me to choose from and also suggested other aspects of the topic I could explore for further information. There is also a modify response option at the bottom of the text that you can select. This will allow you to shorten or lengthen the text, or make it more casual or professional. You can also achieve a similar result by entering a manual prompt.
Another prompt I used was “write a short story about Santa meeting the Easter Bunny in the style of Mary Shelley,” and Gemini provided a chilling tale of Father Nicholas being comforted by the Bunny after his workshop was destroyed by Krampus. Again, I only selected the first of the three drafts available. I realized that after a few other prompt generations, the first one was the only one remaining. Therefore, it’s advisable to examine the options when they are initially generated while you have the chance.
Another prompt I used was “what are the most common CSS grid commands,” which produced a solid list of CSS commands and explanations of their functions. I explored some of the drafts for this response, as well as some included sources. I also performed a few language translations of the term “the dog jumped over the lazy fox” into Spanish, French, and Chinese. The latter provided detailed explanations of different dialects.
A notable aspect of the free version of Google Gemini is its ability to generate images. Not all free AI chatbots possess this capability. This tier of Google Gemini runs on the Gemini Pro LLM, which is considered equivalent to GPT-3.5, the model powering the free version of ChatGPT. However, there are distinctions between the two models. This is where the multimodal aspect of Google Gemini comes into play. You can only access the multimodal functionalities, such as generating images and interacting with video, with the paid ChatGPT Plus tier, which runs the GPT-4 model or later.
Some of the most impressive functions individuals have used with AI involve crafting images using detailed text prompts. Similar to text-to-speech generation, employing words like “create” or “generate” assists the chatbot in understanding you desire a unique result rather than retrieving something from the web. Additionally, strive to minimize spelling and syntax errors in your prompts, although some errors might yield amusing results you don’t mind. You can always edit the prompt text and try again.
Initially, I used the prompt “generate an image of a futuristic car driving through an old mountain road surrounded by nature,” and the images displayed that touch of photorealism that makes AI enjoyable to use. I tested the same prompt in Microsoft’s Copilot chatbot, and while it yielded solid results, they appeared more like drawings or paintings. They were very good, but the cars and the landscape didn’t quite match as well as in the Google Gemini images. Next, I experimented with variations of a “dogs playing poker” prompt, leading to fascinating outcomes. Some were in a modern style. Some were based in the 2000s. Some were based in 2020. I decided to stop once I started getting dog faces on human bodies.
Gemini Advanced is Google’s paid chatbot version, costing $20 per month and offering a host of perks under the Google One AI Premium subscription tier. Gemini Advanced runs the Gemini Ultra 1.0 and 1.5 Pro language models, the latter debuting in February 2024, which excel at understanding longer prompts and the context of previous prompts. An Advanced subscription also increases the context window size to a whopping 1 million characters and grants users early access to new and exclusive features. Google has bundled Gemini Advanced with AI-enabled productivity applications, including Gemini in Gmail, Docs, Sheets, Slides, and Meet. You also get 2TB of storage, Google Photos editing features, 10% back in Google Store rewards, Google Meet premium video calling features, Google Calendar enhanced appointment scheduling, VPN by Google One, dark web monitoring, 3% to 10% back on the Google Store, and additional customer support. There is currently a free two-month trial offer of Google One AI Premium that allows you to test out Gemini Advanced with no risk.
Extensions are an intriguing shortcut within Gemini that allows you to use the chatbot like a supercharged search engine. With the @ mention at the beginning of a prompt, you can direct Gemini to an app or service focus, whether it be Google Flights, Hotels, Maps, Gmail, Drive, Docs, or YouTube. You will then finish your prompt with the task you want accomplished. This could involve finding a file in Google Docs, summarizing a video on YouTube, or researching travel options in Google Flights. Simply ensure that the feature is enabled by accessing Settings > Extensions and activating all the Extensions you want functional.
One criticism that many people have about generative AI in general is data tracking. However, Google is doing its best to put users at ease by including an accessible activity tracker that allows you to turn data tracking on or off at will and to delete your data whenever you desire. Once you have set your preferences, Google will not have access to any data you do not allow. From this area, you have the option to delete individual prompts by hour, by day, or within a custom range. There are also auto-delete settings for every three, 18, or 36 months. However, if you don’t want to deal with all of that, follow the directions below to learn how to turn off your data and delete your activity history.
Select Gemini in the upper-left corner to navigate to the Google Gemini page. Select the Activity time-shaped icon in the bottom-left corner. Select the Gemini Apps Activity option at the top of the page. Choose Turn off or Turn off and delete activity.
